From: Jeff Johnson <jeff.johnson@oss.qualcomm.com>
To: Masi Osmani <mas-i@hotmail.de>,
Christian Lamparter <chunkeey@googlemail.com>
Cc: linux-wireless@vger.kernel.org, ath9k-devel@qca.qualcomm.com
Subject: Re: [PATCH 00/10] carl9170: 802.11n compliance and driver improvements
Date: Thu, 16 Jul 2026 18:09:56 -0700 [thread overview]
Message-ID: <22af4340-5014-47df-9c8f-17ebcb834e64@oss.qualcomm.com> (raw)
In-Reply-To: <AM7PPF5613FA0B6D188CBDBAFC0CE13247A9444A@AM7PPF5613FA0B6.EURP251.PROD.OUTLOOK.COM>
On 3/12/2026 3:37 AM, Masi Osmani wrote:
> The carl9170 driver for Atheros AR9170-based USB WiFi adapters has been
> effectively unmaintained since 2016. While the hardware shipped as
> Draft-N certified, several 802.11n capabilities were never advertised
> to mac80211, diagnostic counters were left as TODO stubs, and some
> hardware features were never wired up.
>
> This series addresses these gaps in 10 independent patches, ordered
> from simple HT capability flags to more involved PHY programming:
>
> Patches 1-3: HT capability corrections
> - Enable SGI_20 (was only SGI_40)
> - Advertise RX STBC (1 spatial stream)
> - Document the SMPS handler (replacing bare TODO)
>
> Patches 4-5: Diagnostic counters
> - Wire up the RX dropped frame counter
> - Track PHY errors via debugfs
>
> Patch 6: TX power calibration
> - Replace hardcoded 18 dBm with per-channel EEPROM values
>
> Patch 7: Recovery hardening
> - Add exponential backoff to prevent restart storms
>
> Patch 8: Antenna diversity
> - Enable fast antenna diversity for 2-chain devices
>
> Patch 9: DFS radar detection
> - Program radar registers, call ieee80211_radar_detected()
>
> Patch 10: Runtime IQ calibration
> - Periodic I/Q recalibration via existing stat_work timer
>
> All patches are individually compile-tested and have been verified
> on real hardware (AVM Fritz!WLAN USB Stick N, AR9001U) running
> kernel 6.18.12. Each patch applies and compiles independently on
> top of the previous ones.
>
> The ath9k PCI driver for the same chipset family serves as the
> primary reference for register programming and capability flags.
>
> Masi Osmani (10):
> carl9170: mac80211: enable Short Guard Interval for 20 MHz
> carl9170: mac80211: advertise RX STBC capability
> carl9170: mac80211: document spatial multiplexing power save handler
> carl9170: rx: wire up dropped frame counter
> carl9170: rx: track PHY errors via debugfs
> carl9170: phy: populate per-channel TX power from EEPROM
> carl9170: main: add exponential restart backoff
> carl9170: phy: enable antenna diversity for 2-chain devices
> carl9170: fw: enable DFS radar detection
> carl9170: phy: add periodic runtime IQ calibration
>
> drivers/net/wireless/ath/carl9170/carl9170.h | 5 +
> drivers/net/wireless/ath/carl9170/debug.c | 2 +
> drivers/net/wireless/ath/carl9170/fw.c | 3 +
> drivers/net/wireless/ath/carl9170/main.c | 45 ++++-
> drivers/net/wireless/ath/carl9170/phy.c | 194 +++++++++++++++++++
> drivers/net/wireless/ath/carl9170/rx.c | 23 ++-
> 6 files changed, 261 insertions(+), 11 deletions(-)
>
> --
> 2.51.0
>
>
In addition to the 10 patches in this series you then posted an additional set
of 2 patches (labeled 11/12 and 12/12) and then another set of patches
(labeled 13/16..16/16). This disjoint labeling makes it hard to track the
patches in patchwork since they aren't all associated to the same series:
https://patchwork.kernel.org/project/linux-wireless/list/?submitter=220736
In addition, the 3 separated disjoint series aren't threaded together in my
e-mail client.
Likewise, posting an update to one patch is not handled well either:
[PATCH v2 07/16] carl9170: main: add exponential restart backoff
When adding to a series or updating one or more patches in a series it is a
best practice to send a new version of the entire series so that the current
state of all patches in the series is captured.
That said, 3 of 16 the patches were ACKed by Christian so I've picked up those
three patches into my 'pending' branch:
carl9170: mac80211: document spatial multiplexing power save handler
carl9170: rx: track PHY errors via debugfs
carl9170: cmd: downgrade transient register I/O errors to wiphy_dbg
Note that I added a wifi: prefix to all to align with the wireless patch style
as previously reminded by Johannes:
<https://lore.kernel.org/linux-wireless/9d093d8a8df188c99b9b652bb56a5c1faecef6c7.camel@sipsolutions.net/>
See:
https://git.kernel.org/pub/scm/linux/kernel/git/ath/ath.git/log/?h=pending
Once you see those have been accepted I suggest you rebase your work on the
ath.git main branch and then post a v3 series containing all of the remaining
patches, incorporating any feedback and experience gained since they were
originally posted.
/jeff
prev parent reply other threads:[~2026-07-17 1:10 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-12 10:37 [PATCH 00/10] carl9170: 802.11n compliance and driver improvements Masi Osmani
2026-03-13 21:14 ` Christian Lamparter
2026-03-31 19:06 ` [PATCH 15/16] carl9170: phy: warm BB reset and same-channel no-op Masi Osmani
2026-07-17 1:09 ` Jeff Johnson [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=22af4340-5014-47df-9c8f-17ebcb834e64@oss.qualcomm.com \
--to=jeff.johnson@oss.qualcomm.com \
--cc=ath9k-devel@qca.qualcomm.com \
--cc=chunkeey@googlemail.com \
--cc=linux-wireless@vger.kernel.org \
--cc=mas-i@hotmail.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox